Nonstationary-Probabilistic Framework to Assess the Water Resources System Vulnerability: Long-Time Robust Plan and Timing

Abstract

Multi-objective water resources systems require long-term operational plans that consider vulnerability to failure under nonstationary conditions due to climate change. The paper presents a probabilistic decision-making framework under nonstationary assumptions to evaluate the robustness of the pre-selected plans and identify the optimal plan using a genetic algorithm approach. The framework incorporates a new metric for maximum allowable time to apply the adaptations and maintaining operational targets without penalties. The framework has four stages for climate exposure identification, production of water supply scenarios, generation of water demand scenarios, and evaluation of system performance. Hydrologic variables considered include precipitation, temperature, and wind speed. The Diyala River Basin in Iraq was used as a case study to test the effectiveness of the framework. Three synthetic pre-selected plans were developed by reducing the demand ratios of the system. Results indicate that current operational rules are robust for flood protection but vulnerable in drought periods. Precipitation changes were dominant in flood and drought management, and temperature and wind speed change effects were significant during drought. Results demonstrated the framework effectiveness to quantify detrimental climate change effects, provide long-term guides for operational planning, and identify the upper limit application time of the adaptation strategies in the system to avert the climate change impact. Framework application suggests an optimal adaptation strategy, robustness examination of the pre-suggested plans, and identify the maximum allowable time for the robust plans. The study represents the first attempt to consider nonstationary hydroclimatic conditions in simulation of supply, demand, and system loss scenarios.